Data SGP
The data sgp package allows users to run analyses of student growth percentiles from the R software environment. The package is free and available for Windows, OSX and Linux. It is recommended that users have at least some familiarity with the basic use of R prior to running SGP analyses.
The lower level functions (that do the calculations) in the SGP package, studentGrowthPercentiles and studentGrowthProjections, require WIDE formatted data whereas higher level function (wrappers for the lower level functions) that provide additional functionality, such as creating graphs of student growth percentiles and plotting students on a growth curve, require LONG formatted data. In general, we strongly recommend using LONG formatted data for all operational analyses as much of the capability of the SGP package is built around it.
Student growth percentiles measure a student’s progress on MCAS relative to other students with similar MCAS performance histories. They are defined for each student as the percentage of students whose scores on the most recent test section fall within a range extending from the 90th percentile to the 0th percentile, with higher numbers representing greater relative growth.
In order to construct a reliable estimate of a student’s current growth percentile, DESE must have a complete and valid record of a student’s MCAS scores in grades 4, 5, 6, and 8, as well as the previous grade-level tests of that student. When the previous year’s MCAS score falls below a minimum threshold for acceptable quality, growth is not reported.
A number of factors can affect the accuracy of student growth percentiles, including estimation errors in both the previous and current test scores used to calculate the growth percentile. These errors are a natural part of the measurement process and would be present even if the test scores were perfect measures of latent achievement traits.
While a student’s growth may be an important indicator of their success, it is important to note that students who achieve high raw scores on previous test sections will also receive a relatively high SGP due to their relative performance. This can lead to a false sense of security for high performing students, as they may notice that their SGP is very close to their highest score. In fact, a large amount of relative growth is required for a student to achieve a higher SGP than their highest score.